Neurodevelopmental disorders, including intellectual disabilities (developmental delay, mental retardation) and autism spectrum disorders (ASDs), are very common, affecting ~3% of the population. Recent evidence has revealed remarkable genetic heterogeneity in these disorders, with a major role for rare, de novo copy number variations (CNVs) or mutations in single genes, suggesting that hundreds of different genes are involved in these disorders. Alternative, complementary approaches to genome-wide association studies (GWAS) are needed for the identification of such rare, de novo mutations to identify the large number of genes implicated. Whole genome cytogenomic arrays are rapidly replacing the karyotype in clinical testing of children with neurodevelopmental disorders, and provide unprecedented resolution to detect submicroscopic, pathogenic CNVs (deletions and duplications) containing small numbers of candidate genes. Tens of thousands of cytogenomic arrays are being done on a clinical basis in the U.S. We propose to leverage this large clinical dataset for identification of rare CNVs associated with neurodevelopmental disorders (hereafter referred to as IDD/ASDs), to develop a Gene Dosage Map for human development and to identify new CNVs and candidate genes responsible for these disorders with the following specific aims: 1) High-throughput CNV and phenotype data mining from large cohorts of patients with IDD/ASDs. A large consortium of clinical cytogenetics laboratories, the International Standard Cytogenomics Array (ISCA) consortium, has been established to share cytogenomic array data in a centralized, public database (dbGaP, NCBI), conservatively estimated at 50-100,000 patient samples/year. Most patients have intellectual disability and ~20% have ASDs. 2) Analysis of the mechanisms and consequences of pathogenic CNVs. Data from pathogenic CNV+ patients (20% of total) will be used to investigate the molecular mechanisms of CNV formation and their phenotypic consequence. Comparison of overlapping regions in the ISCA database vs. normal CNV databases will be used to develop a Human Gene Dosage Map for developmental disorders. 3) Identification of new CNV disorders associated with IDD/ASDs. Two classes of pathogenic CNVs will be investigated further to identify new IDD/ASD disorders: a) recurrent microdeletion/duplication regions with similar phenotypic findings (e.g., our recent discovery that del(17q12) is a cause of autism, especially in males), b) Non-recurrent, pathogenic CNVs will be used for mapping of overlapping cases with similar, distinctive phenotypes (e.g., autism, epilepsy). 4) CNV directed gene discovery for IDD/ASDs. Data from Aim 3 will generate a number of pathogenic CNV regions containing a limited number of candidate genes (1-15) for the IDD/ASD phenotype. To identify the specific genes responsible, comprehensive mutation analysis will be performed in well-phenotyped patients (without pathogenic CNVs) from publicly available sources (e.g., Simons Simplex Collection, Autism Genetics Resource Exchange and the NIMH repositories).